one way spaning, deck plate, local plate design

[ahmd8201]

Hi i am looking at a deck plate that is welded on to trapezoidal members spanning between primary steelwork. I am considering the local design of the deck plate that is 10 mm thick and spans 400mm (ie primary steelwork at 400 mm c/c). The local design load is a 40 kN point load. Obviously considering the deck plate as a continuous beam as a first approximation is not yielding good results. IE it is overutilised UR ~ 2.5. If I consider plate theory, there is no option in roarks for a one way spanning square plate ( assuming 45 degree dispertion, therefore width of plate used is same as span)
If i look at yield line, there is no information on such a case, normally a concentrated load is considered to give a circular yield line. Can anyone point me in the right direction for a document that would have information on this problem, ie a square deck plate supported on two sides, with a central point load applied. would the two options be for yield line either a single failure plane down the middle (in between the support sides, ie parallel) or diagonal failure planes emanating from the point of load application, so this would look like a square plate with diagonal lines intersecting from opposite corners.

If I was to do this on staad as a plate model, I would need to assign spring supports, but I would need to define the displacement and stiffness. the displacement would be the maximum deflection of the plate but how would i calculate the applied spring stiffness?

many thanks

ahmd

 

[Splitrings]

Isn't the spring stiffness the reaction divided by the displacement?

I don't understand what you are modeling but I am guessing you know the displacement of the main steel and you are trying to get your plate model to match these displacements? It may be trial and error. Adjusting each spring value until your plate displacement match you main steel displacements.

 

[ahmd8201]

Hi,

Spring stiffness is reaction divided by displacement. But does that mean that the approach will be iterative. Ie first find out the displacement and the reaction based on simply supported conditions? Calculate the spring stiffness and then replace the simply supports with spring supports and rerun the analysis ?

 

[ukbridge]

I'd just use Pounder's theory, which should be found in Roark as well. To get fudge the one-way spanning problem, use an arbitrarily large value of the aspect ratio a/b so its basically simply suppoirted

 

[Splitrings]

That would be my approach. Adjsut the spring values until the displacement values converge.